The topic of ‘Free Will’ is an uneasy one, it is difficult to understand and difficult to write about. I will begin by discussing how I define free will, and why I frame the debate as I do.
I will continue to explain why we do not have free will and then go on to attack the very concept of free will, arguing that it is, in fact, incoherent and cannot be made sense of. My argument is ‘ontologically neutral’ (the validity of what I say does not depend upon adopting naturalism or supernaturalism), and therefore even if the notion of an immaterial and distinct ‘soul’ is granted, we still do not have free will.
I will close with a brief comment on the implications for religion and ethics.
I have free will if ‘I’ as a conscious agentive ‘self’ can select my next mental states and if ‘I’ as a conscious agentive self can (of my own self volition, immediately and without external influence) change whatever mental states ‘I’ currently have.
If the conditions above are met then it is true that I have free will, and if they are not met then I do not have free will. What I mean by the ‘self’ is just that thing which is what ‘I’ am which is the ‘possessor’, ‘witness’ and ‘experiencer’ of mental states (the conscious ‘me’).
Why should free will require this notion of the self? Traditionally the argument has been about trying to solve the apparently zero-sum game between ‘determinism’ and ‘free will’. If events in the physical world are determined, then no event remains undetermined and all my actions are the result of antecedent events and therefore not free. For this traditional debate ‘free will’ has required to be some kind of action which is not determined by antecedent events. I contend, though, that if the self is not aware of being the agent (‘cause’) of these ‘undetermined’ events, then the self cannot be held sufficiently responsible.
So ‘free will’ cannot only be some mental event which is undetermined – for then randomized mental events might count as free will – it must be a mental event that is determined from the self (which must itself lie outside of determined events).
I think, therefore, that free will is not only about third person objective features of action, it is (principally) about an internal, subjective, and first person ‘ability’ – the ability to choose or change our mental states. Although this distillation of the problem should suffice to justify why I have defined free will as I have there is, in my mind, a much more intuitive and meaningful reason. In order to cash-out our common sense notions of ‘moral responsibility’, ‘praise and blameworthiness’, and ‘normative judgments’ (ethics) we have to have such a definition.
To focus only on the concept of ‘moral responsibility’: when we conceive of ‘moral responsibility’ we typically think that, in some situation, there is ‘someone’ who is responsible for some state of affairs. If I hit a pedestrian on the road, we typically think that ‘I’ am responsible.
The thing which we take to be responsible is some abstract notion of the ‘self’ – ‘I’ did this, and had ‘I’ only been more attentive then this might not have happened; it is ‘my’ fault, ‘I’ am responsible. But it is not enough for the self merely to exist, say, as an epiphenomenon; the self must be the proximal cause of the action. It is, we imagine, ‘I’ who for one reason or another, decided to do something for which I can subsequently be held responsible. The self, in other words, must be some kind of ‘agent’.
In arguing that this self is the agent of its next mental states I make a particular claim about the nature of behavior. I act in the ways that I do because of my reasons for acting that way – reasons being ‘beliefs, desires, intentions, values, goals, etc’ and in short, mental states. It is that I am hungry (some mental state) that I choose to eat (some behavior). If this is true of behavior in general, which I think it is, then if I am to behave in a way which is ‘free’ then it must be that I, as a self am agentive in deciding my next mental states because it is from these mental states, for which ‘I’ am responsible, that I will act.
Having spelled out why I think ‘free will’ means what it means, I will continue now to demonstrate that we do not have it; I will attempt to do this through an illustration. Suppose there is a computer game and in this game the player controls an avatar from the third person. During any part of the game the player has time to select what its avatar will do next (move here, jump there, pick up this object, etc). With a few adjustments I think this is an adequate illustration of what we should commonly think of ‘free will’.
Imagine that the ‘player’ is really what we mean by the ‘self’, that the avatar of the player is like the body of the self. Just as it is from the directives of the player that the avatar behaves as it does, so it is from the directives of the self that the body behaves as it does. In moments when there is a decision to be made about some course of action the player selects from a range of available options what the avatar will do next. And similarly, if we have free will, then the self must also select (or be able to change) mental states which will affect the course of action taken by the body.
Our subjectivity, however, does not reflect the above situation. ‘We’ are never aware of selecting our next mental states; we simply witness them as arising in consciousness. Mental states arise for reasons that are quite inscrutable, or they arise as reactions to our contingent situations. Say I stub my toe, I am never aware that ‘I’ as a conscious agentive self select ‘pain’ to come about, it simply arises and I become the witness, experiencer (and possessor) of this ‘pain’. If I could select which mental state should arise I would definitely be tempted to feel pain, at least, less sharply than I typically do.
I argued that the ‘self’ must be the possessor, witness and experiencer of its mental states, but in order for there to be free will it must also be the agent. Though we clearly possess, witness and experience our mental life, it is not obvious that we select what it should next be like. I imagine that some doubt remains in the minds of my readers, and so, we might attempt a short experiment: if we do have free will, it should be simple – just stop doubting and accept the truth of what I have thus written and come to believe that we do not have free will (pause here and take some time to think about this).
Such a feat cannot be achieved, at least not immediately and from self-volition. The same can be demonstrated with any number of things (believe that you are Michael Jackson; feel happy; start to love your neighbor; hate UNISA; desire food; desire to murder a family member). Our inability to do something so seemingly simple as to select what our next belief will be, or to change beliefs and desires we currently possess, is what I mean when I say that we do not have free will.
Now, what if we came to realize that we desired to murder an innocent person, and what if an intention to stop and obey the law instead simply does not arise? In what way can we, as ‘selves’, be held responsible if we did not select our desires, and could not change them?
There are two ways in which the position can be attacked. The one is to agree with my definition of free will, but demonstrate somehow that we do possess this mystical ability. The other is to disagree with my definition of free will, thereby obviating the common-sense formulations of ‘moral responsibility’. The former contention cannot, I think, get off the ground. It seems to me quite impossible to argue that we have this ‘ability’ when it so clearly is the case that we don’t. The latter, however, is perfectly valid.
The ‘free will’ that I have defined and described is actually impossible to make sense of. My illustration using the ‘player, avatar dichotomy’ is limited in one major way. The situation with the player and avatar is possible because the player already has desires, intentions and beliefs – I already have some idea of how I want to play the game, and from that I can make adequate choices about what I want my avatar to do next. But to export this idea to the concept of the self and the body is impossible.
The self, I argued, is that thing which is meant to bring about mental states. But in selecting mental states, say, deciding to be happy rather than sad, I am exercising some kind of already existing preference. It is already the case, prior to the selection of my happy state, that I desire to be happy – but where does the desire to be happy come from? The role of the self must be to select my desires, but selection of desires is made arbitrary if I do not have existing preferences. The very concept of ‘agentive selection’ in other words presupposes the existence of some preferences.
In order to make sense of this we would have to posit a ‘meta-self’ who would select our initial preferences, and in order to make sense of that there would have to be a ‘meta-meta-self’ for those preferences, ad infinitum. In other words, free will entails an infinite regression of selves in order to maintain its coherency. A further problem is that it would not be clear which one of these selves ‘I’ would really be, nor is it clear whether it is possible that ‘I’ could be all of them at once. If we suppose that there is no ‘meta-self’, and yet that a ‘self’ does select mental states, then its’ selection would be arbitrary and meaningless, rather than agentive. If we suppose that some mental states or preferences do not require selection and are just ‘brute-psychological-facts’, then it means that in the limit we are not actually free, and all of our ‘choices’ are the product, ultimately, of something we could not choose and have no responsibility over.
It is subjectively obvious that we do not have the ‘free will’ that I defined above, but further the very concept of free will is incoherent and cannot really be made sense of. This has implications for our common sense notions of moral responsibility. I had written previously also that, “This hugely undermines 3 main pillars of classical religion. First, original sin cannot be true…Secondly, the idea of cosmic retribution or cosmic reward does not make sense…Thirdly, the problem of evil…[cannot] be dealt with by…invoking free will.” We need, I think, as a species to come together to think differently, therefore, about our common-sense notions of ‘justice’, ‘the good’, and ‘moral responsibility’.
I noted that this was an ontologically neutral argument. This argument is valid if the ‘self’ is realized by a physical process, like materialists would think, and it is valid if the ‘self’ is of some distinct, and perhaps divine, substance. This is less an argument about how the self is realized, and what it is made up of, as it is an argument about the nature of our subjective experience. Our subjective experience is unlike what it would be if we had ‘free will’, and that concept anyway is quite incoherent.
There is more that can be said about free will, and I fear that by now I have exhausted my readers’ attention and will go no further in this article. What I have done is to describe a ‘free will’ that we do not have, but I have not commented on what it is that we do actually have. People often feel as if they are, in some sense, ‘free’, and this is not entirely an illusion. We feel ‘free’ typically because we act (generally) in the ways that we want to.
We desire to act in some way, and therefore act as such (this is what compatibilist think ‘free will’ is – the ability to act as you will. I am questioning our ability to will what we will). I have simply explained that the desire to act in some way is not brought about by the ‘self’. The ‘self’ is merely the witness and the experiencer of those desires, but it is not the agent bringing them to the fore. So I will admit that more can be said on this topic, and perhaps I will tackle those problems in subsequent writings.
In conclusion, ‘free will’ is something that it is thought we typically have, and we use it in order to cash-out our common sense notions of moral responsibility. Not only do we not have the free will we typically think we do, but the concept itself is mysterious and incoherent. Its incoherency is easily demonstrated, as is the fact that we do not have it. This has grave implications for religion. The realization that there is no free will shatters particular doctrines of religion.
Further, if we understand religion as a sort of exposition of the human condition, a kind of explanation of what we are and why we are here, the realization that there is no free will renders so much of religion to be a meager and bad explanation. As a theory, religion just does not stand up to the test of reality. Ethics – and I use the term broadly – needs to be rethought, since much of the popular thinking revolves around a belief in ‘free will’, but in closing I should emphasize that it does not need to be abandoned. To think so would be to take the wrong lesson from this discussion.